[Drmarkf (OP)]

What is the evidence for the number of days of infectivity, not just for persistence of RNA or antigens in body secretions?
(Very much enjoyed the Cambridge Conversations interviews, Chris!)

[chris]

This is very hard to answer with accuracy because we, of course, do not know the status for people who have trivial, sub-clinical or even asymptomatic infection. We also are unclear what happens long-term with immunocompromised patients and patients who have survived a spell in ITU. We are learning rapidly.

Some people clearly clear the infection rapidly and become test negative as soon as their symptoms resolve. That is of course to say that they have less virus on their mucous membranes than the test can detect, although some tests are very sensitive (ours is 8 genome copies). Also, we are assuming that the swab site is representative of the entire body. Other people definitely take longer to clear the infection. Immunosuppressed people seem to shed for longer.

A helpful rule of thumb is to regard a person with symptoms as an infectious person. So when symptoms abate (fever etc) then infectivity falls.

This manuscript produced early in the outbreak from patients in Wuhan is informative: https://www.atsjournals.org/doi/pdf/10.1164/rccm.202003-0524LE

[Drmarkf (OP)]

Thanks, interesting.
So, are there any ‘viability’ assays for coronaviruses? Failing that, are there any parallels with other similar viruses that give some weight to the suggestion that genome-positivity correlates with persistence of infectious virus?
Epidemiological evidence would go some way towards that, but presumably in a locally epidemic situation a household contact might have acquired their infection from other sources?

[chris]

It is possible to do these assays, but they are painstaking and slow. You can do phenotypic assays where you infect cells with samples and count the number of infections produced, which gives you some idea of the ratio of genome templates to viable virus particles.

Virologists also talk about something called the particle:pfu ratio. Put simple, down a microscope you can count viral particles present in a sample. Then you attempt to infect cells with a similar sample and you see how many cells are effectively infected.

What this reveals is that only a proportion of the viral particles you can see down the microscope are actually capable of mounting an infection. The rest are presumably defective in some way, have mutated or damaged genomes, or lack a genome entirely. This could be the consequence of rapid growth, cell stress, immune responses and so on.

Similarly, when we swab a mucous membrane and test for the presence of the viral genetic code, that's merely nucleic acid that we are detecting, not viable virus. It may be, for instance, that the virus particles are all covered in antibody and no longer capable of causing infection. But because we are looking for the nucleic acid template, we still find it.

So at the moment we have no clear data on the quantities and viabilities of shed virus in post-infectious people. Information is emerging, slowly, though...